Cloud-Based Dynamic Warning System Bradley Wentz, Principal Investigator Upper Great Plains Transportation Institute North Dakota State University JUNE 2019 Research Report Final Report 2019-19 • mndot.gov/research
Cloud-Based Dynamic Warning System
Bradley Wentz, Principal InvestigatorUpper Great Plains Transportation Institute North Dakota State University
JUNE 2019
Research ReportFinal Report 2019-19
• mndot.gov/research
To request this document in an alternative format, such as braille or large print, call 651-366-4718 or 1-800-657-3774 (Greater Minnesota) or email your request to [email protected]. Please request at least one week in advance.
Technical Report Documentation Page 1. Report No. 2. 3. Recipients Accession No.
MN/RC 2019-19
4. Title and Subtitle 5. Report Date
Cloud-Based Dynamic Warning System June 2019 6.
7. Author(s) 8. Performing Organization Report No.
Bradley Wentz
Mohammad Smadi 9. Performing Organization Name and Address 10. Project/Task/Work Unit No.
Upper Great Plains Transportation Institute North Dakota State University NDSU Dept. #2880 PO Box 6050 Fargo, ND 58108-6050
11. Contract (C) or Grant (G) No.
(c) 1003323 (wo) 1
13. Type of Report and Period Covered12. Sponsoring Organization Name and Address
Local Road Research Board Minnesota Department of Transportation Office of Research & Innovation 395 John Ireland Boulevard, MS 330 St. Paul, Minnesota 55155-1899
Final Report 14. Sponsoring Agency Code
15. Supplementary Notes
http:// mndot.gov/research/reports/2019/201919.pdf 16. Abstract (Limit: 250 words)
Horizontal curves account for more than 25% of highway fatal crashes and have a crash rate that is three times
that of other highway segments. Transportation agencies employ dynamic warning systems that utilize roadside
speed detection and warning equipment to provide drivers with a warning when they enter a curve at a speed
that might be too high for safe travel. The goal of this project is to develop a system to expand the safety
improvement potential of a dynamic curve speed warning system that can be implemented systemwide to all
reduced speed curves without infrastructure investment. The system emulates a connected vehicle environment
by utilizing a smartphone application to deliver dynamic, speed-based, directional warnings at locations in an
online database. Transportation agencies are able to enter and manage the warning locations within their
jurisdictions in the online warning database through a web-based tool, and it is envisioned that the warning data
would be made available to navigation systems and vehicle manufacturers via the cloud.
17. Document Analysis/Descriptors 18. Availability Statement
Connected vehicles, Highway safety, Cloud computing, Mobile
applications, Highway curves, Speeding
No restrictions. Document available from:
National Technical Information Services,
Alexandria, Virginia 22312
19. Security Class (this report) 20. Security Class (this page) 21. No. of Pages 22. Price
Unclassified Unclassified 51
CLOUD-BASED DYNAMIC WARNING SYSTEM
FINAL REPORT
Prepared by:
Bradley Wentz
Mohammad Smadi
Upper Great Plains Transportation Institute
North Dakota State University
June 2019
Published by:
Minnesota Department of Transportation
Office of Research & Innovation395 John Ireland Boulevard, MS 330
St. Paul, Minnesota 55155-1899
This report represents the results of research conducted by the authors and does not necessarily represent the views or policies
of the Minnesota Local Road Research Board, Minnesota Department of Transportation, or the Upper Great Plains
Transportation Institute at North Dakota State University. This report does not contain a standard or specified technique.
The authors, the Minnesota Local Road Research Board, the Minnesota Department of Transportation, and the Upper Great
Plains Transportation Institute at North Dakota State University do not endorse products or manufacturers. Trade or
manufacturers’ names appear herein solely because they are considered essential to this report.
ACKNOWLEDGMENTS
The research team would like to acknowledge the Minnesota Department of Transportation and
Minnesota Local Research Board, both of which funded the project and made it possible. The team
would also like to thank Otter Tail and Pope counties for their participation in the project and assisting in
testing the system, Otter Tail County Engineer Richard West for serving as the project’s technical liaison,
and Pope County Engineer Brian Giese for participating in field tests.
TABLE OF CONTENTS
CHAPTER 1: Introduction ....................................................................................................................1
1.1 Horizontal Curves ............................................................................................................................... 1
1.1.1 The Problem ................................................................................................................................ 1
1.1.2 Solutions ...................................................................................................................................... 1
1.2 Cloud-Based Dynamic Warning System .............................................................................................. 2
CHAPTER 2: Literature Review ............................................................................................................3
2.1 Concept and Connected Vehicle Application ..................................................................................... 3
2.2 Utilizing Smart Devices (Phones and Tablets) .................................................................................... 3
2.3 Safety Benefits .................................................................................................................................... 4
2.4 Warning Enhancement: Prediction and Vehicle Assisted Actions ...................................................... 4
CHAPTER 3: Cloud-Based Dynamic Warning System ............................................................................5
3.1 Online Warning Database ................................................................................................................... 6
3.2 Warning Application ........................................................................................................................... 7
CHAPTER 4: System Testing .............................................................................................................. 10
4.1 Testing Parameters ........................................................................................................................... 10
4.2 Field Testing ...................................................................................................................................... 10
CHAPTER 5: Benefits and Conclusions ............................................................................................... 12
5.1 Research Benefits ............................................................................................................................. 12
5.1.1 Safety ......................................................................................................................................... 12
5.1.2 Cost Saving ................................................................................................................................ 12
5.2 Conclusions ....................................................................................................................................... 13
REFERENCES .................................................................................................................................... 14
APPENDIX A: Industry Contact Letter
APPENDIX B: National Rural ITS Conference 2018 Presentation
LIST OF FIGURES
Figure 1.1 Examples of horizontal curve warning signs: on the left a static sign, on the right a dynamic
curve speed warning sign. ............................................................................................................................ 2
Figure 3.1 System overview. ......................................................................................................................... 5
Figure 3.2 Dynamic Warning layer in GRIT. .................................................................................................. 6
Figure 3.3 Initial warning display for traveling at or below the advisory speed (left) and above the
advisory speed (right). .................................................................................................................................. 8
Figure 3.4 Secondary warning. ...................................................................................................................... 9
Figure 4.1 A portion of reduced-speed curves in Otter Tail County with County Road 33 test segment
highlighted. ................................................................................................................................................. 11
Figure 4.2 Curve speed warning locations in Pope County. ....................................................................... 11
LIST OF ABBREVIATIONS
API: Application Programming Interface
APK: Android Package
CAV: Connected and Autonomous Vehicles
CV: Connected Vehicle
CSW: Curve Speed Warning
GPS: Global Positioning System
GRIT: Geographical Roadway Inventory Tool
MUTCD: Manual for Uniform Traffic Control Devices
UGPTI: Upper Great Plains Transportation Institute
EXECUTIVE SUMMARY
Horizontal curves account for more than 25% of highway fatal crashes and have a crash rate that is three
times that of other highway segments. To mitigate this problem of vehicles entering horizontal curves at
speeds that may be too high for safe travel, transportation agencies employ a number of solutions
including static reduced speed curve signs and dynamic systems that provide a warning to drivers when
warranted based on their travel speed.
Dynamic curve speed warning systems consist of a detection component to capture vehicle speed and a
warning component in the form of dynamic message signs or flashing beacons with static signs. They
also require power and communications infrastructure and cost roughly $14,000 per site. Although
research shows that dynamic systems are more effective at reducing vehicle travel speeds through
horizontal curves, their deployment is limited to curves with historically high crash numbers due to the
cost involved. The goal of this project is to develop a system to expand the safety improvement
potential of a systemwide dynamic curve speed warning system that can deliver warnings to all reduced
speed curves without infrastructure investment.
The developed system emulates a connected vehicle environment by utilizing a smartphone application
to deliver dynamic, speed-based, directional warnings at locations that have been entered into an online
database. The application captures the vehicle travel speed and direction from the smartphone’s GPS
and utilizes geo-fencing technologies to determine when the vehicle is approaching a reduced speed
location. The application will generate an onscreen and audible warning when warranted based on the
vehicle’s travel speed and advisory speed data from the online warning database. While the smartphone
application is used for proof of concept, the research team sees great value potential in the online
warning database as a way for transportation agencies to enter and maintain reduced speed locations
within their jurisdictions, where the data would be made available to navigation systems and vehicle
manufacturers via the cloud. With the system in place, it could be utilized for situations other than
horizontal curves that require a speed warning, including bumps and cracks due to seasonal changes,
temporary maintenance and construction activities (work zones), special events, and approaching
controlled intersections.
The project deliverables, in addition to this report, are access to the online warning location database
and the smartphone warning application in Android distribution format and the application’s source
code.
1
CHAPTER 1: INTRODUCTION
1.1 HORIZONTAL CURVES
Curves represent a transition element in highway geometric design, while vertical curves are changes in
the slope of the roadway, horizontal curves are changes in the alignment of the road that are necessary
for gradual change in direction when a direct point of intersection is not feasible.
1.1.1 The Problem
Horizontal curves have been shown to be highway segments that require special attention from
transportation agencies due to an overall increase in the rate and severity of crashes occurring at these
locations. In fact, horizontal curves have a crash rate three times that of other highway segments and
account for more than 25% of highway fatal crashes (Albin, et al., 2016).
Horizontal curves occur more frequently on rural two-lane roadways, which is a contributing factor
resulting in the fatality rate on rural roads to be more than twice that of urban roadway segments
(NHTSA, 2007-2016). Two-lane, two-way roadways, predominant in rural Minnesota, experience an
elevated number of roadway departure crashes, resulting in the death of more than 70 people each
year, or nearly 20% of annual roadway fatalities in Minnesota. Furthermore, more than 4,000 people are
injured in these types of crashes every year. Horizontal curves were identified as a contributing
secondary factor in 49% of fatal roadway departure crashes in Minnesota (Leuer, 2015).
1.1.2 Solutions
Transportation agencies have a wide array of countermeasures to mitigate the problem of vehicles
entering horizontal curves at speeds that may be considered too fast for safe travel. Some of these
countermeasures involve the use of basic treatments from the Manual on Uniform Traffic Control
Devices (MUTCD) such as centerline and edge line pavement markings; enhanced basic treatments by
improving the visibility of devices via an increase in sign size and by placing identical signs on both sides
of the roadway; use of reflectors, panels, and chevrons; and rumble strips and other minor roadway
improvements such as the elimination of shoulder drop-off and paving and widening shoulders.
The focus of this research is the mitigation method of utilizing dynamic curve speed warning (CSW)
systems that only provide a warning when warranted to a vehicle that is approaching the curve at a
speed higher than recommended. Dynamic curve speed warning systems consist primarily of two
components: a detection component that measures the vehicle speed via inductive loops or a
nonintrusive method such as radar detection; and a warning component that activates when warranted
and can consist of a static sign with flashing beacons or a dynamic message sign. The system also
requires power and communications for operation bringing the cost to roughly $14,000 per warning
location (USDOT-ITS JPO, 2009).
2
Figure 1.1 Examples of horizontal curve warning signs: on the left a static sign, on the right a dynamic curve
speed warning sign.
As shown in the literature review chapter, dynamic CSW systems are more effective than static signs at
reducing the travel speed of approaching vehicles and reducing the crash rate at the warning location.
However, due to their cost, they are typically deployed at locations with historically high crash rates.
1.2 CLOUD-BASED DYNAMIC WARNING SYSTEM
The goal of the system developed in this project is to extend the safety benefit potential of a dynamic
CSW system to every reduced speed curve in the transportation network taking dynamic warning
onboard the vehicle and removing the infrastructure cost per location associated with traditional
dynamic CSW systems. The system consists of two components: an online database of warning locations
and a warning application.
The warning database contains the coordinates of reduced speed curves. The database is built as an
additional layer within an existing software system, the geographic roadway inventory tool (GRIT)
developed by members of the research team at the Upper Great Plains Transportation Institute at North
Dakota State University. The warning database is available online and is used by the warning application
to determine when and where a warning should be generated.
The warning application is a smartphone application developed on the Android platform. It tracks the
speed, location, and direction of travel of the vehicle; crosschecks that information against data within
the online warning database; and generates warnings as warranted when a vehicle approaches a
location at a speed that may be too high for safe travel.
3
CHAPTER 2: LITERATURE REVIEW
The research team reviewed a variety of materials related to curve speed warning (CSW) under the
following three main areas: information from the Federal Highway Administration (FHWA) on CSW
Connected Vehicle applications and pilot deployment; literature articles related to dynamic warning
applications utilizing both instrumented vehicles and smart devices; and literature articles on the
benefits of dynamic vs static CSW applications. A summary of the literature review is provided below.
2.1 CONCEPT AND CONNECTED VEHICLE APPLICATION
CSW is an application where alerts are provided to the driver who is approaching a curve at a speed that
may be too high for safe travel through that curve (FHWA, 2017). Using equipment such as basic radar
detection sensors or a combination of multiple types of sensors such as radar and environmental
sensors, the infrastructure application component collects available infrastructure and vehicle data. In a
Driver Infrastructure Interface (DII) scenario where the vehicles are not equipped, the driver sees the
alert via a roadway device. If the vehicle is equipped with a Driver Vehicle Interface (DVI), the available
data are processed and the vehicle will display an appropriate advisory message, alert, and/or warning
for the driver on the in-vehicle DVI providing the driver of the connected vehicle with a real time
warning unique to the driver and the vehicle.
Based on the available data and complexity of the algorithm used to evaluate safe speed thresholds, the
message sent to drivers varies. Different operational policies and constraints have been discussed in this
document. These systems have been installed in Michigan, Georgia, and Oregon and are operational
right now.
Michigan Department of Transportation (DOT) Metro Region anticipates improved motorist safety and a
reduction in crashes. In Oregon, the evaluation of the system indicates an effective reduction in the
mean speeds of passenger cars and trucks by around 3 mph for the southbound direction and 2 mph for
the northbound direction. However, this system has its own disadvantages. For example, even if a driver
is traveling at a lower speed than what the roadway signage indicates, due to different conditions such
as weather, it may issue a warning to drivers for high speed which may end up with some confusion for
the driver (Stephens, Timcho, Young, Klein, & Schroeder, 2012). CSW is an especially useful concept for
drivers in unfamiliar surroundings or in adverse weather conditions (Sayer, LeBlanc, Mefford, &
Devonshire, 2007).
There are different methods for warning the driver in an instrumented connected vehicle including
auditory by use of headrest speakers, haptic using brake pulse, and visual using an interface inside the
vehicle with varying degrees of effectiveness (Green, et al., 2008).
2.2 UTILIZING SMART DEVICES (PHONES AND TABLETS)
In a study, researchers utilized a smartphone-based application equipped with Bluetooth Low Energy
(BLE) technology and a head-up display. In this system, messages can be wirelessly communicated from
4
the GPS-equipped smartphone to a receiving unit through BLE technology, and then displayed by head-
up display on the vehicle’s front windshield. Drivers can receive alerts without looking away from their
usual line of sight and the relative accuracy of the model appeared to be high. The authors suggest that
this application can be used for other locations such as work zones and highway-rail grade crossings
(Qin, Zhang, & Wang, 2015).
In a report on connected commercial vehicles, it is shown that CSW could appear on an iPad that is used
as part of the commercial vehicle’s instrument panel. When an alert is warranted, a warning can be
generated as a caution of the upcoming curve or a warning with audible tones (LeBlanc, Bogard, &
Goodsell, 2014).
2.3 SAFETY BENEFITS
A 5% to 7% reduction in crashes was observed in a study in a national demonstration project depending
on crash types and direction of travel when the crash occurred (Hallmark, Hawkins, & Smadi, 2015). A
speed reduction of 3 mph was recorded at a dynamic curve warning sign on I-5 in Myrtle Creek, OR
(Bertini, Monsere, & Nolan, 2006). Average speed reduction of between 1 and 8.8 mph at three low-
volume rural highways in Meeker and McLeod counties MN (Knapp & Robinson, 2012). 68% of drivers
reduced their speed on 5 dynamic curve warning locations on I-5 in Northern California (Tribbett,
McGowen, & Mounce, 2000).
2.4 WARNING ENHANCEMENT: PREDICTION AND VEHICLE ASSISTED ACTIONS
In a summary report for a pilot study that tested warning systems, it was noted that improvements were
needed to reduce unwarranted warnings. Further, driver behavior indicated the application of the
brakes varied among driver groups and situations (LeBlanc, Buonarosa, Blankespoor, & Sayer, 2009).
In a study by (Machiani, Jahangiri, Balali, & Belt, 2017), a model was suggested to improve CSW system
performance through activating the warning using the point of curvature speed instead of approaching
speed. The real world connected vehicle dataset used in this study was a part of the Safety Pilot Model
Deployment from over 2700 vehicles equipped with connected vehicle technology. The results of the
proposed model that considers a driver behavior risky when vehicle approaching speed exceeds the
advisory speed, showed an improvement in false alarms for all the monitoring periods.
5
CHAPTER 3: CLOUD-BASED DYNAMIC WARNING SYSTEM
The Cloud-Based Dynamic Warning system created through this project takes dynamic CSW onboard the
vehicle and negates the need for roadside devices traditionally used for the speed detection and
warning components of a CSW system. The infrastructure expenditure per site that is required for the
deployment of CSW systems limited their deployment to locations with historically high crash rates.
Eliminating the need for roadside devices and their cost allows for the safety improvement potential of
dynamic CSW to be available at every reduced speed curve in the transportation network.
The system consists of an online database of warning locations and a smartphone application that
provides a warning when warranted utilizing the concept of geo-fences to determine when the vehicle
enters a warning location.
Figure 3.1 System overview.
6
3.1 ONLINE WARNING DATABASE
The warning location database component of the system satisfies the following objectives:
Provides transportation agencies with a secure and easy to use method for entering and
maintaining warning locations within their jurisdictions;
Makes the database available online for use in the warning application;
Creates a process for sharing the warning data with third parties such as vehicle and safety
system manufactures, and navigation system developers.
In order to achieve this, the research team utilized a software system already developed by the Upper
Great Plains Transportation Institute called the Geographic Roadway Inventory Tool (GRIT). GRIT is a
web application with a map and graphical user interface, and a SQL database server backbone that
allows agencies to manage an inventory of their transportation assets. Ten Minnesota counties already
use GRIT to manage their data within layers for construction history, owner/load limits, minor
structures, and construction planning. For the purpose of this project a new GRIT layer was created for
curve speed warning.
Figure 3.2 Dynamic Warning layer in GRIT.
7
The warning database contains the following fields:
Geographic Location: the user may enter coordinates for the warning location or select the location on
the map and the coordinates will be populated automatically.
Advisory Type: this dropdown field allows the selection of speed advisory types other than curve speed
warning. Such advisories may include construction, seasonal speed restrictions, etc.
Advisory Speed: the user enters the reduced speed for the warning location.
Highway Legal Speed: this input is for the regular speed limit for that highway section outside of the
warning location.
Sign Facing Direction: the user may select the facing direction for the warning from this dropdown field.
The direction options include the four cardinal directions of North, South, East, and West; and the four
primary inter-cardinal directions of Northeast, Northwest, Southeast, and Southwest.
Enabled: this checkbox indicates whether this is an activated warning location or not.
The GRIT application can be accessed through this link: http://dotsc.ugpti.ndsu.nodak.edu/grit-app-
test/. Warning locations can be viewed by anyone, but in order to add new or edit existing locations a
password protected login to the website is required. Both Otter Tail and Pope counties were provided
with logins in this project.
3.2 WARNING APPLICATION
To prove the concept that dynamic curve speed warning can be achieved onboard the vehicle without
roadside infrastructure, the research team developed an Android smartphone application that utilizes
Google Maps API. The application captures the vehicle’s location, heading, and speed from the
smartphone’s global positioning system (GPS) in real-time which are crosschecked against locations
from the warning database to determine whether a speed warning is warranted.
At initialization, the warning application pulls warning locations from the online database that are within
a 20-mile radius from the current location and creates geo-fences around each point, this process is
repeated every 12 minutes. Each geo-fence serves as a virtual boundary surrounding a warning location,
this allows the application to track when the vehicle enters a warning area and generate warranted
warnings at the appropriate time and location. The warnings are directional in nature based on the
vehicle’s heading and “sign facing direction” field in the warning database to assure that warnings are
generated correctly for the direction of travel.
During development, the warning application underwent iterative parameter refinement and testing to
assure that warnings are only generated when warranted, at the appropriate location, and in a timely
manner that allows for corrective action to be taken by the driver.
8
The warning application has two levels for warnings provided to the driver based on their travel speed.
The initial warning is a caution for approaching a reduced speed curve that displays the curve’s advisory
speed along with the vehicle’s current speed in a green circle if traveling at or below the advisory speed,
or in a yellow circle if exceeding the advisory speed for the curve.
Figure 3.3 Initial warning display for traveling at or below the advisory speed (left) and above the advisory speed
(right).
If speed is not reduced, the secondary warning is triggered which displays a message to slow down along
with the curve’s advisory speed and the vehicle’s travel speed in a red circle, the secondary warning also
includes an audible warning of three successive beeps.
After initialization, the warning application runs in the background and is not required to be the active
screen on the phone to function properly. If there is a different application active on the screen, the
warning application will preempt the active application and display warnings when warranted.
Furthermore, warnings are still generated even if the phone’s screen has timed out and gone into sleep
mode.
It is important to note that the warning application was developed solely for proof of concept, the
application is not intended to be distributed to the general public. Therefore, the application design and
the visual and audible warnings were put in place for the research team to be able to test and
9
demonstrate the system effectively, particularly whether the phone’s GPS would be accurate enough for
warnings to be generated in a timely manner. As such, human factors were not a consideration in the
design of the application, the warning messages, or the testing process.
Figure 3.4 Secondary warning.
The warning application is part of the project deliverables and is provided in Android Package (APK)
distribution format which allows the application to be installed on Android phones. The counties were
provided a copy of the APK file for field testing. Additionally, the warning application source code is also
provided as a project deliverable.
10
CHAPTER 4: SYSTEM TESTING
4.1 TESTING PARAMETERS
Two participating Minnesota counties, Otter Tail and Pope, entered their reduced speed curves into the
warning database utilizing the GRIT interface and that data were used in system testing. The field testing
was conducted under the following parameters:
Condition 1: Approaching the curve at a speed that is at or below the posted advisory speed –
expected outcome: initial warning for approaching a reduced speed curve is displayed with
travel speed in a green circle.
Condition 2: Approaching the curve at a speed that exceeds the posted advisory speed then
slowing down to at or below the advisory speed – expected outcome: initial warning for
approaching a reduced speed curve is displayed with travel speed in a yellow circle, changing to
a green circle after speed reduction occurs.
Condition 3: Approaching the curve at a speed that exceeds the posted advisory speed limit and
not slowing down through the curve – expected outcome: secondary warning is displayed and
the audible warning of three beeps is played in its entirety.
Condition 4: Slowing down to or below the advisory speed after the secondary warning was
triggered – expected outcome: revert back to initial warning and stop playing the audible
warning.
4.2 FIELD TESTING
The initial field test was conducted on 10/15/2018 on Otter Tail County Road 33 which includes a
segment with 15 reduced speed curves. The team drove the segment four times in each direction
resulting in 100% accuracy for the warning application behaving in the expected manner based on the
test conditions outlined above.
The second field test was conducted on 11/29/2018 in Pope County with participation of county
engineer Brian Giese and MnDOT Research videographer Charles Ombati. The team drove on a number
of different Pope County highways and through a large number of reduced speed curves while observing
the warning application’s behavior under the same conditions outlined above. The warning application
behaved as expected with the exception of not providing a warning at one particular curve. After
concluding the test, the research team examined the record for that curve in the warning database and
found that the “sign facing direction” field was entered as South while it should have been East based on
the roadway geometry.
11
Figure 4.1 A portion of reduced-speed curves in Otter Tail County with County Road 33 test segment highlighted.
Figure 4.2 Curve speed warning locations in Pope County.
12
CHAPTER 5: BENEFITS AND CONCLUSIONS
5.1 RESEARCH BENEFITS
The benefits of the Cloud-Based Dynamic Warning system are twofold: improved safety and cost
reduction over traditional dynamic CSW systems.
5.1.1 Safety
The safety benefits come in the form of reduction in travel speed when a driver is traversing a horizontal
curve. Existing in-car safety features such as adaptive cruise control and lane departure warning systems
are fast becoming standard features for many manufacturers. While these features create many safety
improvements, there are some safety concerns with the inability to identify and adjust to hazards such
as reduced speed curves, bumps, and construction and maintenance activities. The online warning
database component of this system could provide a potential basis for transportation agencies to
provide this information to vehicle and safety systems manufacturers. For evaluating the potential
safety benefits, the research team relies on existing literature regarding the benefits of dynamic CSW
systems. These benefits are typically expressed in the amount of reduction in travel speed when utilizing
dynamic vs. static warning or by the reduction of the percentage of vehicles traveling at a certain
threshold above posted or advisory speed, or by performing a crash analysis on the roadway comparing
crash rates before and after a dynamic system is installed.
We hypothesize that the safety benefits of the Cloud-Based Dynamic Warning system will be similar to
levels achieved in traditional roadside dynamic warning systems. The following is a summary of safety
benefits of dynamic speed warning systems from various literature publications.
Crash reduction: 5% to 7% reduction in crashes depending on crash types and direction of travel when
the crash occurred (Hallmark, Hawkins, & Smadi, 2015).
Speed reduction: a speed reduction of 3 mph at a dynamic curve warning sign on I-5 in Myrtle Creek, OR
(Bertini, Monsere, & Nolan, 2006).
Speed reduction: average speed reduction of between 1 and 8.8 mph at three low-volume rural
highways in Meeker and McLeod counties, MN (Knapp & Robinson, 2012).
Speed reduction: 68% of drivers reduced their speed on 5 dynamic curve warning locations on I-5 in
Northern California (Tribbett, McGowen, & Mounce, 2000).
5.1.2 Cost Saving
Roadside dynamic warning systems consist of a detection component to capture vehicle speeds and a
warning component to relay a message to drivers. The detection component may include technologies
such as in-pavement inductive loops, or non-intrusive applications such as radar and cameras. The
warning component may be in the form of a dynamic message sign or a static sign with flashing beacons.
13
To operate these devices, roadside warning systems require power and communication at the warning
location. The cost is roughly $14,000 per warning location (USDOT-ITS JPO, 2009).
Whereas existing dynamic warning systems rely on roadside infrastructure and thus are only deployed
at high crash locations (typically greater than or equal to 10 crashes in a 24-month period and greater
than or equal to 7 crashes in a 12-month period), the proposed system eliminates the need for any
roadside devices and takes the warning system entirely onboard the vehicle, resulting in cost savings
and the advantage of spreading the potential safety benefits systemwide to all reduced speed curves
regardless of crash history. A new warning location in the proposed system requires only an additional
entry into the warning location database.
To express the cost savings benefit, we perform a break-even calculation. The research project cost at
under $80,000 would fund 6 traditional dynamic CSW sites. The cost to maintain the warning application
software and the online warning location database is offset by the cost to maintain the infrastructure of
traditional dynamic CSW systems. Considering that Otter Tail County roadways have over 400 reduced
speed curves, the potential cost-savings benefit is great.
5.2 CONCLUSIONS
The research team developed successfully a system for providing curve speed warnings onboard the
vehicle without utilizing roadside devices and infrastructure. The Cloud-Based Dynamic Warning system
consists of a smartphone warning application and an online warning database managed through a web application.
The purpose of the smartphone warning application is for proof of concept and was utilized for testing
the system; the application will not be available to the general public. However, the online warning
location database is potentially valuable for implementing an in-vehicle CSW system. For managing the
warning locations, the team utilized GRIT and created a layer for the Cloud-Based Dynamic Warning
system. Both Otter Tail and Pope counties entered their reduced speed curves into the system and that
data was utilized for field testing. For any future implementation, the participating transportation
agencies would be responsible for using the GRIT tool to populate the database with the desired
warning locations within their jurisdiction.
CSW is a planned vehicle to infrastructure (V2I) connected vehicle (CV) application. In a CV
environment, the vehicle would take the role of warning a driver who is traversing a curve at an unsafe
speed through different measures such as audible warnings, or haptic feedback through vibrating the
seat or steering wheel. In a connected and autonomous vehicle (CAV) environment, it is perceived that
the vehicle would also take corrective actions automatically, such as slowing down and steering assist to
keep the vehicle on the road, which could greatly improve the safety potential of systems such as CSW.
For these safety systems to function correctly, it is important that transportation agencies are able to
communicate their needs and maintain an inventory of warning locations as they know their roadways
best. The GRIT Dynamic Warning layer can be the tool used for that purpose and for sharing the data
with vehicle and safety system manufacturers and other interested third parties.
14
REFERENCES
Albin, R., Brinkly, V., Cheung, J., Julian, F., Satterfield, C., Stein, W., Donnell, E., et al. (2016). Low-Cost
Treatments for Horizontal Curve Safety 2016. Washington, DC: Federal Highway Administration
Office of Safety.
Bertini, R. L., Monsere, C. M., & Nolan, C. (2006). Field Evaluation of The Myrtle Creek Advanced Curve
Warning System. Salem, OR: Oregon Department of Transportation.
FHWA. (2017). Connected Vehicle Pilot Depolyment Program. Retrieved October 23, 2017, from
Intelligent Transportation Systems - CV Pilot Deployment Program:
https://www.its.dot.gov/pilots/pilots_v2i.htm
Green, P., Sullivan, J., Tsimhoni, O., Oberholtzer, J., Buonarosa, M., Devonshire, J., Schweitzer, J., et al.
(2008). Integrated Vehicle-Based Safety Systems (IVBSS): Human Factors and Driver-Vehicle
Interface (DVI) Summary Report. Springfield, VA: National Highway Traffic Safety Administration
(NHTSA).
Hallmark, S. L., Hawkins, N., & Smadi, O. (2015). Evaluation of Dynamic Speed Feedback Signs on Curves:
A National Demonstration Project. McLean, VA: Federal Highway Administration.
Knapp, K. K., & Robinson, F. (2012). The Vehicle Speed Impacts of a Dynamic Horizontal Curve. St. Paul,
MN: Minnesota Department of Transportation.
LeBlanc, D., Bogard, S. E., & Goodsell, R. (2014). Connected Commercial Vehicles—Integrated Truck
Project Model Deployment Operational Analysis Report. Washington, DC: FHWA.
LeBlanc, D., Buonarosa, M., Blankespoor, A., & Sayer, J. (2009). Integrated Vehicle-Based Safety Systems
(IVBSS) Light Vehicle Extended Pilot Test Summary Report. Washington, DC: National Highway
Traffic Safety Administration (NHTSA).
Leuer, D. (2015). Fatal Ran off the Road Crashes on Rural Two-Lane Two-Way Highways in Minnesota.
Saint Paul, MN: Minnesota Department of Transportation.
Machiani, S., Jahangiri, A., Balali, V., & Belt, C. (2017). Predicting Driver Risky Behavior for Curve Speed
Warning Systems Using Real Field Connected-Vehicle Data. TRB 96th Annual Meeting
Compendium of Papers. Washington, DC: Transportation Research Board.
NHTSA. (2007-2016). Traffic Safety Fact Sheet Rural/Urban Comparsion. Washington, DC: USDOT-
NHTSA.
Qin, X., Zhang, S., & Wang, W. (2015). Advanced Curve-speed Warning System Using an In-Vehicle Head-
Up Display. TRB 94th Annual Meeting Compendium of Papers. Washington, DC: Transportation
Research Board.
15
Sayer, J. R., LeBlanc, D. J., Mefford, M. L., & Devonshire, J. (2007). Field Test Results of a Road Departure
Crash Warning System: Driver Acceptance, Perceived Utility and Willingness to Purchase.
Proceedings of the Fourth International Driving Symposium on Human Factors in Driver
Assessment, Training and Vehicle Design (pp. 77-83). Stevenson, WA: Transportation Research
Board.
Stephens, D. R., Timcho, T. J., Young, E., Klein, R. A., & Schroeder, J. L. (2012). Accelerated Vehicle-to-
Infrastructure (V2I) Safety Applications Concept of Operations Document. McLean, VA: FHWA.
Tribbett, L., McGowen, P., & Mounce, J. (2000). An Evaluation of Dynamic Curve Warning Systems in the
Sacramento River Canyon. Sacramento, CA: California Department of Transportation.
USDOT-ITS JPO. (2009). Unit Cost Component - Dynamic Curve Speed Warning Sign. Retrieved October
25, 2017, from FHWA ITS Costs Database:
http://www.itscosts.its.dot.gov/ITS/benecost.nsf/0/3335052358E73D1385257A7700742C41
APPENDIX A: INDUSTRY CONTACT LETTER
A-1
For part of the project’s Task 2: Literature Review and Automobile Manufacturer Contact, the research
team contacted members of the industry for possible interest in the project and in the online warning
database in particular. The following entities were contacted:
Alliance of Automobile Manufacturers. The Auto Alliance is an advocacy group for the auto
industry representing 70% of all car and light trucks in the US including the BMW Group, Fiat
Chrysler Automobiles, Ford Motor Company, General Motors Company, Jaguar Land Rover,
Mazda, Mercedes-Benz USA, Mitsubishi Motors, Porsche, Toyota, Volkswagen Group of
America and Volvo Car USA.
Honda – Ohio State University Automotive Proving Grounds
HERE Maps
Clemson University International Center for Automotive Research
Easy Mile Driverless Busses
The contact letter is provided in the following page.
A-2
Upper Great Plains Transportation Institute – Advanced Traffic Analysis Center
North Dakota State University Dept 2880
P.O. Box 6050
Fargo, ND 58108-6050
Dear Automobile and Navigation System Manufacturer:
I am writing to inform you of a research project for a vehicle Cloud Based Dynamic Warning
System conducted by our center and sponsored by the Minnesota Local Road Research Board that
could be of benefit to your vehicle/navigation system safety applications.
Over 25% of roadway fatal crashes are associated with a horizontal curve. Research shows that
dynamic curve speed warning systems are more effective than static signs in reducing vehicle
speeds while traversing curves. Due to cost and operational limitations, dynamic warning systems
deployment is limited to locations with historically high crash numbers. The goal of this research
project is to expand the safety benefits of dynamic curve speed warning (CSW) to virtually all
potentially dangerous curves within a jurisdiction by providing on-board dynamic warning to
vehicles entering a horizontal curve at speeds that may be too high for safe travel. Further, safety
benefits can also be achieved at any roadway segment requiring a reduction in travel speed for
additional circumstances such as seasonal factors, special events, and incidents. The research team will be developing a software tool for local jurisdictions to enter and edit their reduced speed
warning locations in an online database that can be provided to automobile and navigation system
manufacturers for use in applications such as Adaptive Cruise Control and Lane Departure
Warning.
The online warning database will at a minimum contain the following fields: coordinates of warning
location; direction of travel the warning is meant for; and the advisory speed for that warning
location.
Please let us know if this project is of interest to you or if you would like to be involved on
discussions regarding the cloud based database and its integration potential with industry.
Contact information: Brad Wentz
Phone: (701)231-7230
Email: [email protected]
APPENDIX B: NATIONAL RURAL ITS CONFERENCE 2018
PRESENTATION
B-1
For the project’s Task 8: Out-of-State Conference Travel, the research team presented the Cloud Based
Dynamic Warning System at the National Rural ITS Conference held October 21-24 2018 in Fort
McDowell, AZ. The presentation was in a session titled: Talking to Each Other: Connected Vehicle and
Infrastructure Applications. The slides from the presentation are provided below.
B-2
B-3
B-4
B-5
B-6
B-7
B-8
B-9
B-10
B-11
B-12
B-13
B-14
B-15
B-16
B-17
B-18
B-19
B-20
B-21
B-22
B-23